SHORTS, voltage across resistor converts to current

[walters]

If "disconnecting" is easily achievable, it's the preferred method. Mostly it's not feasible.

Which Methods would you do?

 

[ahsan_i_h]

hello walters,
Sorry that in older posts, I used the term DVM in place of DMM (digital multimeter) by mistake. Sorry again.

 

[FvM]

Which Methods would you do?

As told, by measuring supply net voltage drop.

 

[walters]

Today at work i found some shorts

On the PCB boards there is 2 regulators in parallel
one outputs 5 volts
the 2nd outputs 3 volts

I was getting 5 volts everywhere , where i was suppost to get it

I was getting 5 volts on stages that were suppost to have 3 volts for +Vcc

How would u start to troubleshooting and what would u do please?

How would u narrow down to which component was causing a short from 5 volts on the 3 volts rail?

1.) I started on a Vcc rail that was suppost to be 3 volts ( even tho it was shorted to 5 volts) , I just used that as a reference point
2.) I used the other probe of my DVM meter to measure the ohms of a component going to +Vcc 5 volts.
3.) the problem was that every component that was tied to the 5 volts +Vcc measure zero ohms and the continuity check was beeping
4.) So the only thing I could do was disconnect each component that was tied to 5 volts +Vcc one by one until I found the short

Was this the best way and how would u have done this please?

 

[walters]

whats an low ohms adapter?

If all components tied up to either the +Vcc 5 volts and +Vcc 3 volts was testing SHORTED

Because The +5 volts rail was shorted to the +3 volt rail

but Only one component was doing that

How would u guys find that component please?

 

[sky_123]

What components were connected to the +5V and +3.3V rails?
And which one was the faulty one?

 

[walters]

These are on the 5volt and 3 volt rails , collector resistors, Pull up resistors, IC chips to power them on , op amps to power them on , or 5volts direct to FETs source and drain pins

But how would u narrow it down?

Some are IC chips, some are FETs, some are pull up resistors, some are collector resistors

How would u track it down?

 

[sky_123]

A real circuit would not have that many devices straddling the +5V and +3.3V rails. This is why I'm asking, what components were there that were across these two
rails, and which one was causing the short (you said one component was doing that, and you desoldered them one by one to find it)?
I'd like to know more about this circuit that has so many components across these two rails, and which component you identified to have a short.

 

[walters]

Yes there is multiple collector resistors , pull resistors, IC chip pins , Transistors pins, FETS pins, some go to the Vcc +5 volts or others go to Vcc +3 volts

I have seen alot of circuits like this , it's nothing special

I have seen alot of transistors , fets, IC chips , collector resistors , pull resistors Tied to Vcc + 5 volt Rails

The circuit is a dual supply , it outputs 5 volts and 3 volts

Nothing special at all

- - - Updated - - -

you said one component was doing that, and you desoldered them one by one to find it)?

Yes how else would u do it? to track down the SHORT or shorted component?

 

[godfreyl]

Yes there is multiple collector resistors , pull resistors, IC chip pins , Transistors pins, FETS pins, some go to the Vcc +5 volts or others go to Vcc +3 volts

Yes, there are many components connected to +5V or +3V. There are not many components connected to +5V and +3V.

I would expect the problem to be with the power supply or regulators. If the 3V is from a linear regulator connected to +5V, I would suspect the 3V regulator is faulty.

 

[FvM]

Yes how else would u do it? to track down the SHORT or shorted component?

I would use the method that I sketched in post #36. It can work for shorts between nets as well.

I'm using a lab power supply, set to a low out voltage, e.g. 0.1 V and current limit 100 mA - 1 A, depending on the involved trace width. Than inject the current between the shorted nets and trace the voltage drop along the circuit.

The idea behind the low output voltage is to protect the circuit if the test source is inadvertently connected to a different net or the short opens unexpectedly during test.

 

[walters]

I'm using a lab power supply, set to a low out voltage, e.g. 0.1 V and current limit 100 mA - 1 A, depending on the involved trace width. Than inject the current between the shorted nets and trace the voltage drop along the circuit.

I thought a Short doesn't have a voltage drop right?

Can you inject 1 amp of current? won't that damage the transistors,fets and IC chips?

- - - Updated - - -

I'm using a lab power supply, set to a low out voltage, e.g. 0.1 V and current limit 100 mA - 1 A, depending on the involved trace width. Than inject the current between the shorted nets and trace the voltage drop along the circuit.

I thought a Short doesn't have a voltage drop right?

Can you inject 1 amp of current? won't that damage the transistors,fets and IC chips?

- - - Updated - - -

Than inject the current between the shorted nets and trace the voltage drop along the circuit.

Do I inject it at the power supply output , after the regulator , or do I inject it at the Vcc and ground rails of a faulty stage?

 

[FvM]

I thought a Short doesn't have a voltage drop right?

The short itself may have a voltage drop, but usually it hasn't. The voltage drop occurs along the traces between the short and the point where you connect the power supply. So it indicates the location of the short.

Can you inject 1 amp of current? won't that damage the transistors,fets and IC chips?

A low voltage of 0.1V doesn't forward bias semiconductors. There may be other components that can be damaged with 1A, in this case you should reduce the current limit setting.

 

[walters]

A low voltage of 0.1V doesn't forward bias semiconductors. There may be other components that can be damaged with 1A, in this case you should reduce the current limit setting.

So should I but the low voltage at 0.7volts so it can turn on the semiconductor components?

How do you know what to set the current limit to ?

If I don't know where the SHORT is , where Do I Inject the lab power supply at in the circuit?

 

[FvM]

My suggestion was to set the voltage clearly below 0.7 V.

The current should be large enough to see voltage drops. It also depends on the resolution of your voltmeter. If 100 mA are enough, it's fine, otherwise higher currents should be used.

I would normally inject the current at the regular power supply entry respectively behind voltage any voltage regulators.

 

[walters]

The current should be large enough to see voltage drops.

So you will see voltage drops except where the short is?

I'm just not sure what to do after you hook up the lab power supply to 0.1 volt and current limit to 1Amp '

You hook up the lab powers supply to the "input" of the regulator of the circuits powers supply?

Then you probe each stage of the circuit measuring voltage drops?

How does that lead you to the SHORT?

 

[FvM]

I suggest to sketch a supply network with a short using pencil and paper. Assume realistic trace resistances and calculate the expectable voltage drops.

A PCB with 35 µm copper plating (1 oz) has about 0.5 mOhm sheet resistance. A 0.5 mm wide supply trace has 1 mOhm resistance per mm, so you get 1 mV per mm voltage drop at 1A.

Supplying 0.1V to a voltage regulator input usually results in zero output voltage and not current. I suggested to connect the voltage source behind the regulator (= to the regulator output). Depending on the regulator behaviour, it may be also possible to supply a higher current limited voltage to the regulator input.

 

[walters]

Ok I connect it to the output of the regulator

0.1 volt at 1 amp is injected into the circuits Rails/planes

What do i do now with the DVM meter? and how does this lead me to the SHORT please?

 

[walters]

FvM help, did u get the message?

 

[FvM]

I think, I explained the method detailed enough to clarify, where voltage drops can be observed. According to ohms law, it's along those traces that carry the short circuit current. Assuming a star supply net topology, it's only one of many traces. In case of a meshed supply net, there will be still a systematical voltage gradient towards the short location.